Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Distributed Monitoring of Historical Web Page Progression for Predictive Presentation Purposes

IP.com Disclosure Number: IPCOM000015900D
Original Publication Date: 2002-Jun-15
Included in the Prior Art Database: 2003-Jun-21
Document File: 3 page(s) / 49K

Publishing Venue

IBM

Abstract

Distributed Monitoring of Historical Web Page Progression for Predictive Presentation Purposes Disclosed are two fundamental ideas: (1) web page histories across a broad selection of users to perform predictive caching at the desktop or at the internet service provider; (2) the idea of using aggregated sequential web page histories across a broad selection of users to perform user prompting of possible "interest pages" of relevance to the current topic. Over the past few years, the distribution of content over the world wide web has fundamentally changed the way that people exchange information. And, the explosion of business models has led to many new and innovative ways of delivering heightened information to the desktop. Mechanisms that will speed the delivery and efficiency, and methods that both speed the download process and bring germane pages to the user's attention would be highly desirable.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 47% of the total text.

Page 1 of 3

Distributed Monitoring of Historical Web Page Progression for Predictive

Presentation Purposes

Disclosed are two fundamental ideas: (1) web page histories across a broad selection of users to perform predictive caching at the desktop or at the internet service provider; (2) the idea of using aggregated sequential web page histories across a broad selection of users to perform user prompting of possible "interest pages" of relevance to the current topic.

Over the past few years, the distribution of content over the world wide web has fundamentally changed the way that people exchange information. And, the explosion of business models has led to many new and innovative ways of delivering heightened information to the desktop. Mechanisms that will speed the delivery and efficiency, and methods that both speed the download process and bring germane pages to the user's attention would be highly desirable.

In this invention, proposed are new methods developed to perform exactly these functions. They require information to be sent back from the browser to a data collection service, presumably presented via a user-selectable setting on the browser. Detailed considerations of privacy may be discussed and managed beyond the scope of the architectural layout presented here. The process will run as follows, with definitions of key terms given below:

Interval(RealTime) == a moving window of time in which historical usage patterns are considered relevant. Note that this value is independent of any html tags which may be embedded in the documents themselves, pertaining to expiration. Sequential(Patterns) == the sequence of web pages hit by a given web user. Hits(Threshold) == Number of hits on a given website before a predictive model is considered for presentation to users. Hit(Percentage) == Percentage of aggregate hits, representing how many users went from a given "page A" to given "page B." Hit(ThresholdPercentage) == Minimum Hit(Percentage) before a given web page is offered for presentation to users.

(1) Each time a user hits a web page, the information is sent back anonymously, from either an agent running on the browser, or an agent running at the ISP, to a central collection point. Note that if the agent runs on the ISP using aggregated hits occurring from all clients, then privacy concerns become even more greatly mitigated.

(2) Logic at that collection point looks at trends in two fashions: Interval(RealTime) and Sequential(Patterns). The variable Interval(RealTime) comprises the moving window where subsequent hits and Sequential(Patterns) are considered relevant. This may be a window as small as a few minutes, or as large as 48 hours or more, although any option could be considered for the art. The second parameter Sequential(Patterns), is

1

Page 2 of 3

the series of hits arising from a single browser instance. For instance, a given user's sequence may be defined as "page A," "page B," "page C," "page D." Another user's sequence may...